Date of Award

Spring 2008

Document Type


Degree Name

Doctor of Philosophy in Industrial Engineering - (Ph.D.)


Industrial and Manufacturing Engineering

First Advisor

Sanchoy K. Das

Second Advisor

Layek Abdel-Malek

Third Advisor

George Hanna Abdou

Fourth Advisor

Athanassios K. Bladikas

Fifth Advisor

Rene Cordero


Evaluation of supply chain performance is often complicated by the various interrelationships that exist within the network of suppliers. Currently many supply chain metrics cannot be analytically determined. Instead, metrics are derived from monitoring historical data, which is commonly referred to as Supply Chain Analytics. With these analytics it is possible to answer questions such as: What is the inventory cost distribution across the chain? What is the actual inventory turnover ratio? What is the cost of demand changes to individual suppliers? However, this approach requires a significant amount of historical data which must be continuously extracted from the associated Enterprise Resources Planning (ERP) system.

In this dissertation models are developed for evaluating two Supply Chain metrics, as an alternative to the use of Supply Chain Analytics. First, inventory costs are estimated by supplier in a deterministic (Q , R, δ )2 supply chain. In this arrangement each part has two sequential reorder (R) inventory locations: (i) on the output side of the seller and (ii) on the input side of the buyer. In most cases the inventory policies are not synchronized and as a result the inventory behavior is not easily characterized and tends to exhibit long cycles. This is primarily due to the difference in production rates ( δ), production batch sizes, and the selection of supply order quantities (Q) for logistics convenience. The (Q , R, δ )2 model that is developed is an extension of the joint economic lot size (JELS) model first proposed by Banerjee (1986). JELS is derived as a compromise between the seller's and the buyer's economic lot sizes and therefore attempts to synchronize the supply policy. The (Q , R, δ )2 model is an approximation since it approximates the average inventory behavior across a range of supply cycles. Several supply relationships are considered by capturing the inventory behavior for each supplier in that relationship. For several case studies the joint inventory cost for a supply pair tends to be a stepped convex function.

Second, a measure is derived for responsiveness of a supply chain as a function of the expected annual cost of making inventory and production capacity adjustments to account for a series of significant demand change events. Modern supply chains are expected to use changes in production capacity (as opposed to inventory) to react to significant demand changes. Significant demand changes are defined as shifts in market conditions that cannot be buffered by finished product inventory alone and require adjustments in the supply policy. These changes could involve a ± 25% change in the uniform demand level. The research question is what these costs are and how they are being shared within the network of suppliers. The developed measure is applicable in a multi-product supply chain and considers both demand correlations and resource commonality.

Finally, the behavior of the two developed metrics is studied as a function of key supply chain parameters (e.g., reorder levels, batch sizes, and demand rate changes). A deterministic simulation model and program was developed for this purpose.